Determination of Curie Point Depth and Heat Flow Using Airborne Magnetic Data over the Kom-Ombo and Nuqra Basins, Southern Eastern Desert, Egypt

The Kom-Ombo and Nuqra basins in southern Egypt have recently been discovered as potential hydrocarbon basins. The lack of information about the geothermal gradient and heat flow in the study area gives importance to studying the heat flow and the geothermal gradient. Several studies were carried out to investigate the geothermal analyses of the northwestern desert, as well as the west and east of the Nile River, using density, compressive wave velocity, and bottom hole temperature (BHT) measured from deep oil wells. This research relies on spectral analysis of airborne magnetic survey data in the Kom-Ombo and Nuqra basins in order to estimate the geothermal gradient based on calculating the depth to the bottom of the magnetic source that caused the occurrence of these magnetic deviations. This depth is equal to the CPD, at which the material loses its magnetic polarisation. This method is fast and gives satisfactory results. Usually, it can be applied as a reconnaissance technique for geothermal exploration targets due to the abundance of magnetic data. The depth of the top (Zt) and centroid (Z0) of the magnetic source bodies was calculated for the 32 windows representing the study area using spectral analysis of airborne magnetic data. The curie-isotherm depth, geothermal gradient, and heat flow maps were constructed for the study area. The results showed that the CPD in the study area ranges from 13 km to 20 km. The heat flow map values range from 69 to 109 mW/m2, with an average of about 80 mW/m2. The calculated heat flow values in the assigned areas (A, B, C, and D) of the study area are considered to have high heat flow values, reaching 109 mW/m2. On the other hand, the heat flow values in the other parts range from 70 to 85 mW/m2. Since heat flow plays an essential role in the maturation of organic matter, it is recommended that hydrocarbon accumulations be located in places with high heat flow values, while deep drilling of hydrocarbon wells is recommended in places with low to moderate heat flow values.

Investigation of Curie Point Depth in Sulu Ultrahigh-Pressure Metamorphic Belt,Eastern China

The Curie point depth of continental crust can reflect the regional tectonic pattern and geothermal structures. Analysis of magnetism is an efficient way to obtain the Curie point depth on a regional scale. This study systematically investigated the Curie point depth of Sulu （苏鲁） ultrahigh pressure （UHP） metamorphic belt （33°40＇N to 36°20＇N and 118°E to 120°E, ca. 60 000 km^2）, eastern China using aeromagnetic data. The results show that the Curie point depth of the Sulu region varies from 18.5 to 27 km. The shallowest Curie point depth （ca. 18.5 km） is located in Subei （苏北） subsidence, where the estimated temperature gradient value is about 31.35℃/km, which is comparable with the measured value of 30 ℃/km. In addition, a two-dimensional numerical solution of the heat conduction was used to calculate the temperature field to a depth of 30 km along the profile from Tancheng （郯城） to Lianshui （涟水） with a length of 139 km. The steady state model solved using the finite element method shows that the temperature around the Curie point depth is about 585.36 ℃, which is close to the Curie temperature （580℃） of magnetite at atmospheric pressure. These results provide new insights into the tectonic and continuous thermal structures of the Sulu UHP metamorphic belt.

Determination of Curie Point Depth, Heat Flow and Geothermal Gradient from High Resolution Aeromagnetic Data around Lamurde Area, Adamawa State, North-Eastern Nigeria

Analysis of high resolution of aeromagnetic data was carried out over Lamurde, Adamawa state north-eastern Nigeria to determine the Curie point depth (CPD), heat flow and geothermal gradient. The aeromagnetic data used for this work was obtained at Nigerian geological survey agency, the total magnetic intensity was processed to produce the residual magnetic map which was divided into 4 overlapping blocks, each block was subjected to spectral analyses to obtain depths to the top boundary and centroid, while depth to bottom of the magnetic sources was calculated using empirical formula. The depths values obtained were then used to assess the CPD, heat flow and geothermal gradient in the area. The result shows that the CPD varies between 9.62 and 10.92 km with an average of 10.45 k, the heat flow varies between 150.73 and 132.78 mWm−20⋅°C−1 with an average of 139.12 mWm−20⋅°C−1 and the geothermal gradient in the study area varies between 12.16 and 15.67 °C/km with an average of 13.39 °C/km. In view of the above results, the high heat flow may be responsible for maturation of hydrocarbon in Benue Trough as well as responsible for the lead Zinc Mineralization. Again by implication, Lamurde area can be a good area for geothermal reservoir exploration for an alternative source for power generation.

Estimation of Curie Point Depth (CPD) across the Pan African Belt in Northern Cameroon from Aeromagnetic Data

The depth to the Curie isothermal point (CPD) was estimated using aeromagnetic data from northern Cameroon. The CPD was estimated using a two-dimensional power density spectral analysis of nineteen 30 × 30 km blocks over a series of Neoproterozoic lithologies associated with the Pan African orogeny. The results show that there are two main regions characterized by the CPD analysis: 1) Shallow depths (~6.37 - 10.09 km) which are located in several regions including the northern portions of the study area (toward the regions of Hina Marbak, Gawel and northern Moutouroua), the southeast (Kaele region), the south (Guider) and the southwest (Mayo Oulo), and 2) deeper depths (~10.68 - 13.72 km) are located in the northeast (Mindif) and southwest (southern Moutouroua, Guider and Bossoum). The shallow depths can be related to two tectonic regimes: 1) the West-Central African Rift System with northeast-trending strike-slip faults emanating from the Gulf of Guinea and 2) the Cameroon Volcanic Line. However, the ultimate source of these shallow regions is interpreted to be related to the Cameroon Volcanic Line based on low seismic velocities imaged by recent broadband seismic studies which are concentrated along the northeast-trending strike-slip faults. An additional finding using the CPD depths, a Curie isothermal temperature of 580&#176C and a one-dimensional heat flow model, was heat-flow values ranging from 105.68 to 227.63 mW/m2, which are above average global heat flow values and are therefore indicative of potential geothermal resources.

Depth to the bottom of magnetic layer in South America and its relationship to Curie isotherm,Moho depth and seismicity behavior

We have estimated the DBML（depth to the bottom of the magnetic layer） in South America from the inversion of magnetic anomaly data extracted from the EMAG2 grid. The results show that the DBML values, interpreted as the Curie isotherm, vary between -10 and -60 km. The deepest values（〉-45） are mainly observed forming two anomalous zones in the central part of the Andes Cordillera. To the east of the Andes, in most of the stable cratonic area of South America, intermediate values（between -25 and-45 km） are predominant. The shallowest values（〈-25 km） are present in northwestern corner of South America, southern Patagonia, and in a few sectors to the east of the Andes Cordillera. Based on these results, we estimated the heat flow variations along the study area and found a very good correlation with the DBML. Also striking is the observation that the thermal anomalies of low heat flow are closely related to segments of flat subduction, where the presence of a cold and thick subducting oceanic slab beneath the continent, with a virtual absence of hot mantle wedge, leads to a decrease in the heat transfer from the deeper parts of the system.After comparing our results with the Moho depths reported by other authors, we have found that the Curie isotherm is deeper than Moho in most of the South American Platform（northward to -20°S）, which is located in the stable cratonic area at the east of the Andes. This is evidence that the lithospheric mantle here is magnetic and contributes to the long wavelength magnetic signal. Also, our results support the hypothesis that the Curie isotherm may be acting as a boundary above which most of the crustal seismicity is concentrated. Below this boundary the occurrence of seismic events decreases dramatically.

Estimation of Curie Depth of Subei Basin by Magnetic Anomaly

The Curie point depth is the depth at which the ferromagnetic mineral loses paramagnetism,which corresponds to a temperature of 550-580 degrees Celsius.This parameter is of great significance for constraining the thermal state of the lithosphere and can be used as an important reference for the evaluation of the prospects for dry hot rock exploration.We use aeromagnetic anomaly radial power spectrum based on the fractal magnetization model to calculate the Curie depth in the Subei Basin.We use the conventional method to preprocess the aeromagnetic data and process aeromagnetic data with selecting the appropriate wave number,domain,window size and fractal parameters to calculate the depth of the bottom of the magnetic source in the Subei Basin,which is approximated as Curie depth.The Curie depth of Subei basin ranges from 18 km to 37 km,with an average Curie depth of 23 km.The Curie depth of Baoying(20 km)and Jianhu(22 km)area is relatively shallow,and that of the surrounding area is relatively deep.In most areas,the Curie depth corresponds to the terrestrial heat flow.

基于PointNet++的机器人抓取姿态估计

为解决在无约束、部分遮挡的场景下对部分遮挡的物体生成可靠抓取姿态的问题,基于PointNet++网络改进了一种抓取姿态估计算法,该算法可直接从目标点云中生成二指夹具的抓取姿态。由于该算法降低了抓取姿态的维度,将抓取的7自由度问题转变成4自由度问题处理,从而简化学习的过程加快了学习速度。实验结果表明:该算法在无约束、部分遮挡的场景中,能够生成有效的抓取姿态,且较Contact-GraspNet算法成功抓取率提升了约12%,能够应用于家用机器人的抓取任务。

Enhanced 3D Point Cloud Reconstruction for Light Field Microscopy Using U-Net-Based Convolutional Neural Networks

This article describes a novel approach for enhancing the three-dimensional(3D)point cloud reconstruction for light field microscopy(LFM)using U-net architecture-based fully convolutional neural network(CNN).Since the directional view of the LFM is limited,noise and artifacts make it difficult to reconstruct the exact shape of 3D point clouds.The existing methods suffer from these problems due to the self-occlusion of the model.This manuscript proposes a deep fusion learning(DL)method that combines a 3D CNN with a U-Net-based model as a feature extractor.The sub-aperture images obtained from the light field microscopy are aligned to form a light field data cube for preprocessing.A multi-stream 3D CNNs and U-net architecture are applied to obtain the depth feature fromthe directional sub-aperture LF data cube.For the enhancement of the depthmap,dual iteration-based weighted median filtering(WMF)is used to reduce surface noise and enhance the accuracy of the reconstruction.Generating a 3D point cloud involves combining two key elements:the enhanced depth map and the central view of the light field image.The proposed method is validated using synthesized Heidelberg Collaboratory for Image Processing(HCI)and real-world LFM datasets.The results are compared with different state-of-the-art methods.The structural similarity index(SSIM)gain for boxes,cotton,pillow,and pens are 0.9760,0.9806,0.9940,and 0.9907,respectively.Moreover,the discrete entropy(DE)value for LFM depth maps exhibited better performance than other existing methods.

基于空间传播的多视图三维重建

针对多视图三维重建任务中点云完整性欠佳的问题,提出一种基于空间传播的多视图深度估计网络(SPMVSNet)。引入空间传播思想用于复杂条件下的稠密点云重建,并分别设计基于空间传播的混合深度假设策略和空间感知优化模块。混合深度假设策略采用由粗糙到精细的深度推理方式,将深度估计视为多标签分类任务,对正则化概率体执行交叉熵损失以约束代价体,从而避免回归方法过拟合和收敛速度过慢的问题。空间感知优化模块从包含高级语义特征表示的特征图中获得引导,在进行置信度检查后采用卷积空间传播网络,通过构建亲和矩阵来细化最终的深度图。同时,为解决大多数方法存在的对不满足多视图一致性的不可靠区域重建质量较低的问题,进一步结合注意力机制设计具有样本自适应能力的动态特征提取网络,用于增强模型的局部感知能力。实验结果表明,在DTU数据集上,SP-MVSNet的重建完整性相比于CVP-MVSNet提升32.8%,整体质量提升11.4%。在Tanks and Temples基准和Blended MVS数据集上,SP-MVSNet的表现也优于大多数已知方法,取得了良好的三维重建效果。

大孔隙参数对斜坡非均匀渗流与稳定性的影响

为揭示降雨条件下大孔隙参数对斜坡水分非均匀运移与稳定性的影响,基于两域模型与稳定系数场原理,建立降雨入渗下斜坡非均匀渗流与稳定性求解模型,并借助COMSOL Multiphysics多物理场有限元平台,编制相应的模型求解程序,通过大孔隙土柱降雨试验验证数值模型的合理性,对比均匀流与非均匀流条件下斜坡体积含水率和点稳定系数,分析大孔隙参数(大孔隙占比ω_(f)、两域导水系数之比μ、大孔隙经验参数r_(w))对斜坡渗流场及稳定系数场的影响规律。结果表明:相比不考虑大孔隙,考虑大孔隙时的基质域和大孔隙域表层体积含水率分别增长7.7%和降低5.1%,入渗深度分别增长83.3%和150.0%;边坡浅层失稳面积增大3.9%。基质域和大孔隙域入渗深度均随大孔隙占比ω_(f)的增大而减小;随着大孔隙域与基质域饱和渗透系数之比μ增大,两者入渗深度变化趋势相反,即μ越大,基质域入渗深度越小,大孔隙域反之;两者与经验参数r_(w)无显著关系。至降雨结束,基质域表层土体体积含水率已达最大值;大孔隙域则随着ω_(f)和μ的增大而增大,但几乎不受经验参数r_(w)的影响。非均匀流条件下,边坡水分交换沿着剖面从上往下分为负交换区、正交换区和无交换区,水分交换平衡深度与基质域入渗深度变化趋势一致;水分交换负交换区与正交换区的深度均存在一个峰值,并随大孔隙占比ω_(f)的增大而减小,随着μ和r_(w)的增大而增大。不同参数取值下,边坡均为浅层失稳破坏,ω_(f)和μ越大,失稳层深度越大,表层点稳定系数越小,因此大孔隙不利于边坡稳定。

基于伪点云特征增强的多模态三维目标检测方法

环境感知是自动驾驶汽车落地的关键技术之一,它对于提高自动驾驶汽车的安全性和可靠性至关重要.三维目标检测是其中的一项核心任务,旨在识别和定位三维空间中的物体,为后续决策提供重要的信息.点云和图像是该任务最常用的输入数据,点云由三维空间中不规则分布的点组成,而图像则是由二维空间上规则分布的像素组成.因此,点云和图像之间难以进行有效的融合.而伪点云作为一种点云表征的图像信息,近几年受到了该领域学者的广泛关注.现阶段基于伪点云的三维目标检测方法还存在伪点云特征提取粗糙和相应感兴趣区域(Region-of-Intersts,RoI)特征表征能力差的问题.本文针对上述问题开展研究,分别提出细粒度注意力卷积和多尺度分组稀疏卷积.细粒度注意力卷积将规则图像处理中常用的深度可分离卷积引入不规则点云的处理流程,并在此基础上嵌入通道和分组注意力机制,进行精细的特征提取,增强伪点云特征;多尺度分组稀疏卷积将格网池化后的Ro I特征分组,进行差异化特征学习,获取不同尺度的Ro I特征,增强伪点云Ro I格网特征的表征能力.基于此,本文在SFD(Sparse Fuse Dense)网络的伪点云特征提取流程中引入细粒度注意力卷积,同时在其伪点云Ro I特征学习流程中引入多尺度分组稀疏卷积,构建SFD++多模态三维目标检测网络.在权威KITTI自动驾驶数据集上的实验结果表明,SFD++每秒可以处理8.33帧数据,其精度在简单、中等和困难的三维汽车检测上达到95.74%、88.80%和86.04%,比次优SFD的精度高出0.15%、0.84%和0.58%.除此之外,一系列消融和补充实验结果验证了所提出卷积的有效性和相关参数设置的合理性.

基于锚点的快速三维手部关键点检测算法

在人机协作任务中,手部关键点检测为机械臂提供目标点坐标,A2J(Anchor-to-Joint)是具有代表性的一种利用锚点进行关键点检测的方法。A2J以深度图为输入,可实现较好的检测效果,但对全局特征获取能力不足。文中设计了全局-局部特征融合模块(Global-Local Feature Fusion,GLFF)对骨干网络浅层和深层的特征进行融合。为了提升检测速度,文中将A2J的骨干网络替换为ShuffleNetv2并对其进行改造,用5×5深度可分离卷积替换3×3深度可分离卷积,增大感受野,有效提升了骨干网络对全局特征的提取能力。文中在锚点权重估计分支引入高效通道注意力模块(Efficient Channel Attention,ECA),提升了网络对重要锚点的关注度。在主流数据集ICVL和NYU上进行的训练和测试结果表明,相比于A2J,文中所提方法的平均误差分别降低了0.09 mm和0.15 mm。在GTX1080Ti显卡上实现了151 frame·s^(-1)的检测速率,满足人机协作任务对于实时性的要求。

肾俞穴穴性及用法新探

肾俞穴作为肾脏之气输注于背腰部的腧穴,其属于足太阳膀胱经范畴,在临床上广泛运用于生殖泌尿系统等疾病的治疗。该穴位于人体背部足太阳膀胱经直行支脉循行处,且为足少阴经脉之标部所在,故具有特殊的穴性。但当前对于肾俞穴穴性及运用缺乏深入研究。该文基于经脉、标本、四海等理论探究了其穴性,并通过古籍溯源及临床实践探究其疗效发挥与刺灸方法、穴位配伍、针刺深度等关系。结果表明,不同的刺灸方法、穴位配伍、针刺深度能激发肾俞穴不同的穴位疗效,因此,只有明辨肾俞穴穴性并辅予合适的刺灸方法、穴位配伍及针刺深度,才可更好地发挥肾俞穴的治疗效果。

基于三维点云的光栅三维显示景深优化方法

对三维显示内容进行深度压缩处理,是一种针对裸眼光栅三维显示器景深范围不足问题的有效解决方法,但常用的压缩方法会使场景中的主体不可避免地产生几何形变。本文面向密集视点光栅三维显示,提出了一种基于三维点云的景深优化方法,根据三维场景的双目视差图重建出三维点云,将场景主体点云分割,仅对超出显示器景深范围的主体进行深度位置调整,保持场景主体的几何结构不变,从而实现三维场景深度的整体压缩。对本文方法进行关于视觉体验的多项统计实验,对于每个参与者,本文方法处理后结果得票率均超过77%;对于每个实验场景,本文方法得票率均超过80%。实验结果证明了本文方法对观众主观感受有明显提升,可以在压缩场景整体深度的同时保证场景主体没有形变,让观众在观看三维场景时没有不自然感且对场景主体保留较强的深度感知。

多模态融合的三维语义分割算法研究

如何高效提取稠密感知的图像特征信息以及真实三维感知的点云特征信息并充分利用其各自优势进行信息互补是提升三维目标识别的关键。本文提出了一种图像和点云融合的多模态框架用于三维语义分割任务。图像与点云特征提取分支相互独立,设计深度估计融合网络用于图像分支,将稠密感知的图像语义信息与真值显式监督的深度特征信息有效融合,对点云的无序及稀疏性进行补偿。并改进体素特征提取方法,减少点云体素化带来的信息损失。图像、点云分支提取多尺度特征后通过动态特征融合模块提升网络对关键特征的提取能力,更有效的获取全局特征。同时本文提出点级的多模态融合数据增强策略,提升样本多样性的同时有效缓解样本不均衡问题。在Pandaset公开数据集上进行对比实验,本文的多模态融合框架展现出更优的性能和更强的鲁棒性,尤其在小样本小目标上性能提升更为明显。

基于RealSense仓库货物体积计算方法的研究

企业仓库中,查看可用空间量是调配货物过程中的重要环节,而对于固定的仓库,它的总容量是固定的,所以可以通过算出仓库中现有的货物量,然后两个相减即可得到剩余可用空间量。对于计算仓库货物的体积,首先使用RealSense深度相机获取货物场景的初始点云。由于获取的初始点云数据往往存在噪声,所以在进行货物体积计算时需要进行一些预处理。预处理过后得到了货物的点云,然后利用投影法和积分法对其进行计算测量。针对仓库货物堆的大小不一问题,提出了两种方案:第一种,对于小货物堆(体积小于2立方米的货物),使用相机从上往下拍,获取货物的顶面和地面,然后通过统计分析法、投影与积分的结合得到货物体积;第二种,对于大货物堆(体积大于等于2立方米的货物)不易获取货物的顶面,使用两个有一定距离的相机获取货物堆的两个侧面,然后对方案一的方法进行变形得到货物体积。实验结果表明:该方式测得的货物体积与实际的货物体积相差不超过3%,满足目前企业仓库的需求。

基于深度相机的大尺寸船体曲板测量技术研究

文中设计了一种基于深度相机的船体曲板测量系统.该系统将深度相机与伺服驱动结合,使用计算机控制采集过程,获取指定位置的多组小范围点云.采用一种超体素聚类改进的区域生长分割方法从原始点云数据中提取曲板表面点云,通过相机安装的相对位置和伺服系统的返回参数值拼接完整的船体曲板三维点云.使用该系统在3 000 mm×3 000 mm范围测量船体曲板,所得非边缘点云平均误差为0.61 mm,符合预期精度要求.

小角度缠绕复合材料护岸管桩抗弯性能试验与设计

本文提出一种新型复合材料护岸管桩。该护岸管桩由纤维缠绕工艺制备的复合材料缠绕管制成。通过四点弯曲试验,得到试件的破坏模式以及荷载-位移关系,以此分析缠绕角度±55°、±15°以及是否浇筑混凝土对护岸管桩抗弯性能影响。结果表明:纤维缠绕角度由55°减小为15°与浇筑混凝土都能极大提高复合材料护岸管桩的刚度和抗弯承载力。在此基础上进行复合材料护岸管桩嵌固端深度计算公式推导,并给出护岸管桩的设计流程以及应用建议。这为之后复合材料护岸管桩应用于护岸工程提供了参考。

基于门控循环深度范围预测网络的多视角重建

针对三维重建技术难以处理高分辨率图像、重建后的点云图精度低、边界模糊的问题,本文提出基于门控循环单元的多阶段多尺度动态深度范围预测网络模型。首先,利用曲率引导的动态尺度卷积网络作为特征提取模块,通过计算图像上多个尺度的表面法曲率,得到图像最优像素的特征信息;然后,将精细的特征信息与一种新的深度范围估计模块相结合,动态估计下阶段的深度范围假设,从而更好的合并邻域像素的信息,实现参考图像和源图像之间的精确匹配。本文网络与其他10多种方法进行了比较,在DTU数据集上,整体性能比第2的网络提高2.2%。在Tank&Temple数据集上,Lighthouse、M60和Panther等场景的重建表现都有大幅提升。同时,本文进行了对比和消融实验,实验结果证明本文提出的动态深度预测网络,减小内存消耗的同时,显著提高了重建后点云图的精度和完整度。

基于深度传感器的多视角点云配准研究

为了解决大尺寸对称物体在多视角配准过程中出现的误匹配点对和累计误差问题,提出了一种基于深度传感器的多视角点云配准算法。首先,使用深度传感器获取目标物体不同视角下的多片点云并进行预处理,对物体单侧相邻点云采用超四点快速鲁棒匹配算法(Super 4-points congruent sets,Super4PCS)进行粗配准,利用改进的点到平面ICP算法去除误匹配点对并进行精配准,之后将左右两部分的点云拼接,从而获取完整的三维点云模型。最后,针对多视角配准出现的累计误差问题,提出了一种全局优化方法从而减少累计误差。实验结果证明所提方法可以精准地完成多视角点云配准,获得准确的三维点云模型。